HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 2006-0229v1 25/7/1635    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hosen, N. Articles by Weissman, I. L. Search for Related Content PubMed PubMed Citation Articles by Hosen, N. Articles by Weissman, I. L.

CANCER STEM CELLS

Bmi-1-Green Fluorescent Protein-Knock-In Mice Reveal the Dynamic Regulation of Bmi-1 Expression in Normal and Leukemic Hematopoietic Cells

Departments of ^aPathology and
^bDevelopmental Biology, Stanford University School Of Medicine, Stanford, California, USA;
^cStanford Institute for Stem Cell Biology and Regenerative Medicine, Palo Alto, California, USA

Key Words. Bmi-1 • Green fluorescent protein • Hematopoietic stem cells • Leukemia

Correspondence: Naoki Hosen, M.D., Ph.D., Department of Cancer Stem Cell Biology, Osaka University, Graduate School of Medicine, 1-7, Yamada-Oka, Suita, Osaka 565-0871, Japan. Telephone: +81-6-6879-3676; Fax: +81-6-6879-3677; e-mail: hnaoki{at}imed3.med.osaka-u.ac.jp

Received April 17, 2006; accepted for publication March 20, 2007.
First published online in STEM CELLS EXPRESS   March 29, 2007.

	ABSTRACT

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
The ability to self-renew is essential for all kinds of stem cells regardless of tissue type. One of the best candidate genes involved in conferring self-renewal capacity is Bmi-1, which has been proven to be essential for the maintenance of both normal adult hematopoietic and leukemia stem cells, as well as adult neural stem cells. To investigate the possible role of Bmi-1 in other cell types that also self-renew, we generated Bmi-1-green fluorescent protein (GFP)-knock-in mice, in which GFP was expressed under the endogenous transcriptional regulatory elements of the Bmi-1 gene. Using these targeted reporter mice, we demonstrated that Bmi-1 is expressed in hematopoietic stem cells (HSCs) at its highest levels and downregulated upon commitment to differentiation. An in vivo reconstitution assay revealed that the frequency of HSCs was 1/16 in Bmi-1^highc-kit⁺lin^–Sca-1⁺ bone marrow (BM) cells and 1/49 in Bmi-1^highlin^– BM cells, suggesting that Bmi-1 may serve as a marker for normal HSCs. In murine leukemia models induced by P210BCR/ABL or TEL/PDGFßR + AML1/ETO, Bmi-1 was not overexpressed in leukemic HSCs, despite the increase in the HSC numbers. Bmi-1 was expressed at its highest levels in undifferentiated leukemia cells. Furthermore, in several other nonhematopoietic tissues, cells could be separated into distinct subpopulations with differential Bmi-1 expression. Thus, these mice allow for the isolation of viable Bmi-1-expressing cells and have the potential to become a useful tool for understanding the role of Bmi-1 in normal and cancer stem cells in multiple tissue types.

Disclosure of potential conflicts of interest is found at the end of this article.

	INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
The prospective isolation of stem cells is the most important step to functionally understand stem cells in any kind of tissue. The strategy that is most commonly used for stem cell isolation is purification based on the expression of a combination of cell surface markers. For example, adult murine hematopoietic stem cells (HSCs) express the surface markers c-kit, Sca-1, Thy1.1, and CD150 but do not express Flk2, CD34, or a panel of antigens associated with lineage commitment [1–4]. An alternative method is to use targeted reporters in which a marker such as green fluorescent protein (GFP) is expressed under a stem cell-specific promoter. For example, Nestin-GFP [5] mice are used to take advantage of the observation that Nestin is expressed in early neural stem and progenitor cells to isolate an enriched population of neural stem and progenitor cells. This approach is unique in that cell surface proteins, as well as intracellular reporter genes, can be used as stem cell markers. Thus, as genes that are highly expressed in one or more types of stem cells are identified, GFP-reporter mice can be generated using their regulatory elements to potentially isolate other tissue-specific stem cells.

Recent evidence has suggested that Bmi-1 is an excellent candidate to specifically mark multiple types of stem cells. The Bmi-1 gene belongs to the polycomb gene family and is a component of the polycomb repressive complex 1 (PRC1) that is implicated in the stable maintenance of gene repression [6]. Extensive analysis of Bmi-1-deficient mice revealed the essential role of this molecule in the self-renewal of both HSCs [7] and NSCs [8]. Furthermore, the Bmi-1 gene is also essential for the self-renewal of at least some leukemic stem cells [9]. In human and mouse hematopoietic cells, Bmi-1 expression is highest in immature hematopoietic cells [10, 11]. In neural cells, Bmi-1 expression is detected in neurospheres [8] and in the external germinal layer of the mouse cerebellum [12]. It has also been reported that Bmi-1 is expressed in epithelial stem cells [13]. Moreover, Bmi-1 is overexpressed in several types of cancer [14]. These results led us to hypothesize that Bmi-1 might be expressed in many kinds of normal and cancer stem cells.

To begin to test this hypothesis, we generated targeted reporter mice in which the endogenous Bmi-1 gene was replaced through homologous recombination with GFP. In this study, we demonstrated that high Bmi-1 expression levels accurately mark HSCs, and we showed that Bmi-1 expression correlates with cell immaturity in leukemia models. Finally, we present data to suggest that high Bmi-1 expression levels have the potential to mark stem or progenitor cells in nonhematopoietic tissues.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
Generation of Bmi-1^GFP/+ Mice and Bmi-1^Bmi-1-GFP/+ Mice
To generate Bmi-1^GFP/+ mice, 1.5- and 5.5-kilobase fragments flanking the Bmi-1 open reading frame were cloned by polymerase chain reaction (PCR) from genomic DNA of R1 embryonic stem (ES) cells as 5' and 3' homologous arms to prepare the targeting vector. The enhanced green fluorescent protein (EGFP) gene was fused in-frame to the ATG at the Bmi-1 translation start site in the 5' homologous arm. The resulting 5' arm and 3' arm were inserted into 5' and 3' multiple-cloning sites of the pKS-TK-Neo-LoxP vector (a kind gift from Dr. Rajewsky, Harvard Medical School) to construct the targeting vector (Fig. 1A). Targeted ES cell clones were obtained by homologous recombination in R1 ES cells. The neomycin resistance gene was deleted by transient transfection of the Cre recombinase expression vector pPAC-Cre (a kind gift from Dr. Yagi, Osaka University). Targeted clones were identified by Southern blot analysis with the flanking 5' and 3' probes shown in Figure 1B. The targeted ES cells were injected into blastocysts of C57BL/6 mice to obtain chimeric founders. The obtained Bmi-1^GFP/+ heterozygotes were backcrossed with C57BL/6/Ka–Thy1.1, CD45.2 mice. The targeting vector for the Bmi-1^Bmi-1-GFP/+ mice was generated as shown in Figure 1D. EGFP was fused in-frame with exon 10 in the 5' homologous arm, resulting in production of a Bmi-1-GFP fusion protein. Mice were bred and maintained at Stanford University under animal ethics guidelines.

View larger version (28K): [in this window] [in a new window]   Figure 1. Generation of the Bmi-1GFP/+ and Bmi-1Bmi-1-GFP/+ mice. (A, D): The mouse Bmi-1 locus and the targeting construct for the Bmi-1GFP/+ (A) or Bmi-1Bmi-1-GFP/+ (D) mouse. (B, E): Southern blotting to confirm the homologous recombination. (C, F): Bone marrow cells from the Bmi-1GFP/+ (C) or Bmi-1Bmi-1-GFP/+ (F) mouse were separated into GFP+ and GFPlo/– cells and Bmi-1 expression levels were measured by quantitative reverse transcription-polymerase chain reaction. Abbreviations: EGFP, enhanced green fluorescent protein; GFP, green fluorescent protein; kb, kilobase.

Immunophenotype Analysis
Single-cell suspensions from bone marrow, spleen, and fetal liver were 150 mM NH₄Cl, 10 mM KHCO₃-treated for 3 minutes on ice to lyse erythrocytes, washed with phosphate-buffered saline containing 2% fetal calf serum, incubated with 20 mg/ml rat IgG for 20 minutes to prevent nonspecific antibody binding, and stained with the indicated fluorochrome-conjugated antibodies for 30 minutes on ice. Cells were then washed and resuspended in 1 µg/ml propidium iodide. Analysis was performed on a Diva-FACSVantage instrument equipped with a 488 nm argon laser and 599 nm dye laser (Becton, Dickinson and Company, San Diego, http://www.bd.com) or a FACSAria instrument (BD Biosciences, San Diego, http://www.bdbiosciences.com).

Brain cells from postnatal (P)0 mice were prepared by papain digestion. Brains were digested in Earle's balanced salt solution buffer containing 0.36% glucose, 1 mM EDTA, 26 mM NaHCO₃, 2 U/ml papain (Worthington, Lakewood, NJ, http://www.worthington-biochem.com), 0.25 mg/ml L-cysteine (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), and 150 U/ml DNase (Worthington) at 37°C for 1 hour. Digested cells were filtered through a 70-µm nylon mesh, washed with Dulbecco's modified Eagle's medium, and then subjected to staining for fluorescence-activated cell sorting (FACS) analysis. Biotin-conjugated-anti-CD133 antibody and streptavidin-Cy7PE (eBiosciences, San Diego, CA, http://www.ebioscience.com) were used for staining.

Liver nonparenchymal cells were prepared by a two-step perfusion method [15] with collagenase D (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) followed by digestion of the nonparenchymal fraction with papain. CD45.2 (AL1–4A2)-Alexa594, Ter119-Cy5PE (eBiosciences), CD49f-PE (BD Pharmingen, San Diego, http://www.bdbiosciences.com/pharmingen), c-met-biotin (R&D Systems Inc., Minneapolis, http://www.rndsystems.com), and streptavidin-Cy7PE were used for staining.

Single-cell suspensions from testis were prepared by enzymatic digestion as reported by Kubota et al. [16]. Cells were stained with CD49f-PE, c-kit (2B8)-APC, CD45.2-Alexa594, and Ter119-Cy5PE (eBiosciences).

In Vivo Reconstitution Analysis
C57BL/6-CD45.1 mice were lethally irradiated (950 cGy) and injected retro-orbitally with FACS-purified bone marrow cells from Bmi-1^GFP/+ mice (CD45.2) along with a radioprotective dose of 2.5 x 10⁵ syngenic whole bone marrow cells. Peripheral blood from reconstituted mice was analyzed at 4, 8, and 12 weeks post-transplantation using B220 and CD3 markers for B and T cells, respectively, and Mac-1 marker for myeloid cells. Mice were maintained on antibiotics for at least 6 weeks after irradiation. For limiting dilution analysis, calculation of the HSC frequency was performed using bioinformatics software at the Walter and Eliza Hall Institute of Medical Research (http://bioinf.wehi.edu.au/software/limdil/index.html).

Induction of Leukemia Using Bone Marrow Cells from Bmi-1^GFP/+ Mice
MSCV-p210BCR/ABL (a kind gift from Dr. Warren Pear, University of Pennsylvania) or MSCV-TEL/PDGFßR-IRES-AML1/ETO (a kind gift from Dr. Michael Thomasson, Washington University) retrovirus was produced by transient transfection of each viral construct to Phoenix-E producer cell lines (a kind gift from Dr. Gary Nolan, Stanford University) by CaPO₄ precipitation. Bmi-1^GFP/+ mice were treated with 150 mg/kg 5-fluorouracil by intraperitoneal injection. Five days after injection, bone marrow (BM) cells were harvested and cultured in Iscove's modified Dulbecco's medium containing 5% fetal bovine serum, SCF (10 ng/ml; R&D Systems Inc.), Flt3L (10 ng/ml), and IL11 (10 ng/ml) for 24 hours. Cultured cells were infected with MSCV-p210 BCR/ABL or MSCV-TEL/PDGFßR-IRES-AML1/ETO retrovirus in the presence of 4 µg/ml polybrene (Sigma-Aldrich) for 24 hours. Transduced cells were collected and transplanted into lethally irradiated C57BL/6-Thy1.2-CD45.1 recipient mice (2–5 x 10⁵ cells per mouse).

Immunohistochemistry
Immunofluorescence was performed on frozen sections. Sections were blocked prior to staining using the M.O.M. blocking kit (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com) and stained with anti-GFP antibody conjugated with Alexa488 (Molecular Probes Inc., Carlsbad, CA, http://probes.invitrogen.com) for 30 minutes at room temperature. Nuclei were labeled with Hoechst 33342 (Molecular Probes). Immunofluorescent labeling was analyzed using a Nikon Eclipse E800 fluorescent microscope (Tokyo, http://www.nikon.com) or a Leica SP2 AOBS confocal microscope (Allendale, NJ, http://www.leica.com).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
Generation of Bmi-1-GFP-Knock-In Mice
To produce targeted Bmi-1 reporter mice, we generated two different lines of GFP-knock-in mice. In the first line, GFP was used to replace exon 2 of the Bmi-1 gene through homologous recombination (Fig. 1A, 1B). The targeted allele results in transcription of GFP mRNA instead of Bmi-1. Although Bmi-1^GFP/+ mice were heterozygous for the Bmi-1 allele, Bmi-1^GFP/+ mice were phenotypically indistinguishable with respect to survival, hematopoietic cellularity, and lineage composition from wild-type littermates (data not shown), as previously reported in the Bmi-1^+/– mice [17]. To examine whether GFP expression level correlated with Bmi-1 expression in Bmi-1^GFP/+ mice, BM cells were separated into GFP^high and GFP^lo/– populations by FACS sorting and subjected to quantitative RT-PCR analysis for Bmi-1 expression. Bmi-1 expression level in the GFP^high population was approximately 10-fold higher than that of the GFP^lo/– population (Fig. 1C).

In the second line of knock-in mice (Bmi-1^Bmi-1-GFP/+), the GFP gene was fused to exon 10 of the Bmi-1 gene in-frame through homologous recombination, and the targeted allele results in transcription and translation of Bmi-1-GFP fusion gene (Fig. 1D, 1E). Fusion of GFP to Bmi-1 did not appear to affect Bmi-1 function, since the heterozygous Bmi-1^Bmi-1-GFP/+ mice, as well as the homozygous Bmi-1^{Bmi-1-GFP/Bmi-1-GFP} mice, were indistinguishable from littermate wild-type controls in terms of survival, hematopoietic cellularity, and lineage composition (data not shown). The GFP expression level correlated well with the Bmi-1 expression (Fig. 1F).

Bmi-1^GFP Is Expressed at Its Highest Levels in Long-Term HSCs and Downregulated Along with Cell Differentiation in Normal Hematopoietic Cells
Bmi-1 expression in murine and human hematopoietic cells has been previously analyzed by nonquantitative or semiquantitative RT-PCR. These reports showed that Bmi-1 was expressed in HSC-enriched stem and progenitor populations in mice [7, 9, 10] and humans [11]. However, the results obtained from these studies could not distinguish whether all cells within the population express moderate levels of Bmi-1 or whether rare cells within the population express high levels of Bmi-1. The Bmi-1-GFP-knock-in mice allowed us to address this issue. We examined the GFP fluorescence of different hematopoietic cell populations (Fig. 2A) from bone marrow and spleen. The levels of Bmi-1 expression were quantified as mean fluorescence of GFP expression by these cells. First, we analyzed Bmi-1^GFP/+ mice (Fig. 2B, 2C). Bmi-1 was expressed at significantly higher levels in HSC-enriched cells (KLS [c-kit⁺lin^–Sca-1⁺]) [1] compared with other populations. All cells in the HSC population expressed Bmi-1, but there was some heterogeneity in the expression levels. Bmi-1 expression levels in common lymphoid progenitors (lin^–/Thy1.1^–/IL7R⁺/c-kit^int/Sca-1^int) [18] were lower than those in HSCs, but higher than those in mature T and B cells. Similarly, myeloid progenitor cells, including common myeloid progenitors (lin^–/c-kit⁺/Sca-1^–/CD34⁺/FcR^lo), granulocyte monocyte progenitors (lin^–/c-kit⁺/Sca-1^–/CD34⁺/FcR^high), and megakaryocyte erythrocyte progenitors (lin^–/c-kit⁺/Sca-1^–/CD34^–/FcR^lo) [19], expressed Bmi-1 at lower levels than HSCs but at higher levels than mature granulocytes. Next, we analyzed BM and spleen cells from the Bmi-1^Bmi-1-GFP/+ mice in similar fashion. The results were almost identical to those from the Bmi-1^GFP/+ mice, although GFP expression levels were approximately 3–4-fold lower than those in the Bmi-1^GFP/+ mice (Fig. 2D, 2E). Therefore, we used the Bmi-1^GFP/+ mice for further analysis, since GFP signals were easier to distinguish from background than in the Bmi-1^Bmi-1-GFP/+ mice.

View larger version (38K): [in this window] [in a new window]   Figure 2. GFP expression level is highest in HSCs and downregulated along with differentiation. (A): The staining profiles from the Bmi-1GFP/+ mouse and the gates for HSCs, CLP, CMP, GMP, and MEP. (B, D): GFP expression levels in subpopulations of BM and spleen cells from the Bmi-1GFP/+ (B) or Bmi-1Bmi-1-GFP/+ (D) mice. (C, E): The MFI value of each subpopulation from the Bmi-1GFP/+ (C) or Bmi-1Bmi-1-GFP/+ (E) mice is shown as a bar graph (n = 3). Abbreviations: B, B cell; BM, bone marrow; CLP, common lymphoid progenitor; CMP, common myeloid progenitor; FSC, forward scatter; GFP, green fluorescent protein; GMP, granulocyte monocyte progenitor; Gr, granulocyte; HSC, hematopoietic stem cell; MEP, megakaryocyte erythrocyte progenitor; MFI, mean fluorescence intensity; MP, myeloid progenitor; N.D., not done; T, T cell.

View larger version (49K): [in this window] [in a new window]   Figure 3. Long-term (LT)-HSCs express Bmi-1 at the highest levels in normal hematopoietic cells. (A, B): The GFP fluorescence intensities in subpopulations of KLS (c-kit+lin–Sca-1+) cells in BM of the Bmi-1GFP/+ mouse were analyzed on gated LT-HSCs (Flk2–CD34–KLS [A], CD150+CD34–KLS [B]) or other populations. The mean fluorescence of GFP in each population is shown as bar graph (n = 3). Abbreviations: FSC, forward scatter; GFP, green fluorescent protein; MFI, mean fluorescence intensity; PI, propidium iodide.

View larger version (38K): [in this window] [in a new window]   Figure 4. In vivo reconstitution analysis of the Bmi-1high KLS (A–D) and Bmi-1highlin– (E, F) bone marrow cells from the Bmi-1GFP/+ mice. (A): The expression of long-term HSC markers in the GFPhigh and GFPlow KLS BM cells. (B, C): The analysis of peripheral blood chimerism at 12 weeks post-transplant with 50 GFPhigh or GFPlow KLS cells. Shown are the representative fluorescence-activated cell sorting plots (B) and the chimerism in the mice transplanted with GFPhigh KLS cells (n = 5) (C). (D): Limiting dilution analysis with the GFPhigh KLS cells. (E): The expression of HSC markers in the Bmi-1high lin– BM cells. (F): Limiting dilution analysis with the Bmi-1high lin– cells. Abbreviations: GFP, green fluorescent protein; HSC, hematopoietic stem cell; KLS, c-kit+lin–Sca-1+.

We next examined whether functional HSCs could be enriched by the expression level of the Bmi-1 and lineage markers. Five, 50, or 200 lin^–GFP^high BM cells from Bmi-1^GFP/+ mice were FACS-sorted and transplanted into lethally irradiated mice. The lin^–GFP^high BM cells represented 0.18% of whole BM cells and 25.7% of them were c-kit⁺Sca-1⁺Flk2^– cells (Fig. 4E). Engraftment at 12 weeks post-transplant was detected in 0 of 5, 4 of 6, and 5 of 5 of the recipients injected with 5, 50, and 200 cells, respectively. From these results, the frequency of HSCs in the lin^–GFP^high population was estimated at 1 of 49 cells (Fig. 4F). Collectively, these results show that the Bmi-1 expression level is a reliable marker for functional HSCs.

Bmi-1 Expression in the Chronic Myeloid Leukemia Induced by P210BCR/ABL
It was recently reported that Bmi-1 is overexpressed in myeloid dysplastic syndrome (MDS) and acute myeloid leukemia (AML), particularly in the immature FAB M0 subtype [20]. In MDS, Bmi-1 expression levels correlated with disease progression [21]. However, it is still unclear which cells express Bmi-1 in these hematological malignancies. There are two possible models: the first suggests that Bmi-1 expression is abnormally upregulated by oncogenic mutations. The second predicts that high levels of expression of Bmi-1 only reflect the accumulation of immature cells.

To address this issue, we examined Bmi-1 expression in two different mouse leukemia models that have been previously established and well-characterized: (a) chronic myeloid leukemia (CML) induced by P210BCR/ABL [22], and (b) AML induced by the combination of TEL/PDGFßR and AML1/ETO [23]. Bone marrow cells from 5-fluorouracil-treated Bmi-1^GFP/+ mice were retrovirally transduced with P210 BCR/ABL or TEL/PDGFßR-IRES-AML1/ETO and transplanted into lethally irradiated recipients (Figs. 5A, 6A).

View larger version (36K): [in this window] [in a new window]   Figure 5. Bmi-1 expression in the mouse CML-like disease. (A): The scheme of the experimental design. (B): The characteristics of CML-like disease. May-Giemsa staining of PB is shown at a magnification of x400. (C): Absolute numbers of HSCs (Flk2– KLS cells) in femurs and tibias of the CML mice and the mice transplanted with untransduced bone marrow (BM) cells (normal Bmi-1GFP/+ mice). (D): Representative staining profiles of the BM cells from the CML mice and the normal Bmi-1GFP/+ mice. Bmi-1 expression levels in each subpopulation are shown as histograms. (E): Lineage-marker negative (lin–) BM cells of the CML mice were separated into the GFPhigh, GFPint, GFPlo/– cells and compared with respect to the frequencies of immature KLS leukemia cells. Abbreviations: CML, chronic myeloid leukemia; FSC, forward scatter; GFP, green fluorescent protein; HSC, hematopoietic stem cell; MP, myeloid progenitor; PB, peripheral blood.

View larger version (42K): [in this window] [in a new window]   Figure 6. Bmi-1 expression in mouse myeloid leukemia induced by TPAE. (A): The scheme of the experimental designs. (B): The characteristics of TPAE leukemia mice. May-Giemsa staining of PB and H&E staining of BM are shown at a magnification of x400. (C): The absolute numbers of HSCs (Flk2– KLS cells) in BM of the leukemia mice and the mice transplanted with untransduced BM cells (normal Bmi-1GFP/+ mouse). (D): Representative staining profiles of the BM cells from the TPAE leukemia mice and the normal Bmi-1GFP/+ mice. Bmi-1 expression levels in each subpopulation are shown as histograms. (E): Lineage-marker negative (lin–) BM cells of the leukemia mice were separated into GFPhigh, GFPint, and GFPlo/– cells and compared with respect to the frequencies of immature KLS leukemia cells. Abbreviations: BM, bone marrow; 5-FU, 5-fluorouracil; GFP, green fluorescent protein; HSC, hematopoietic stem cell; MP, myeloid progenitor; PB, peripheral blood; TPAE, TEL/PDGFßR + AML1/ ETO.

In the CML mice, the total number of HSCs (Flk2^– KLS cells) in BM cells harvested from bilateral femurs and tibias were approximately twofold more than in mice transplanted with untransduced Bmi-1^GFP/+ BM cells (normal Bmi-1^GFP/+ mice) (6,632 ± 1,308 cells vs. 3,290 ± 729 cells, respectively; n = 3) (Fig. 5C). We examined the Bmi-1 expression level in the HSC population. The Bmi-1 expression level of HSCs in CML mice was similar to that of normal Bmi-1^GFP/+ BM cells (Fig. 5D). In addition, the Bmi-1 expression level in FLK2⁺ KLS cells and myeloid progenitor populations in CML mice was also similar to that of the normal Bmi-1^GFP/+ mice (Fig. 5D). These results demonstrate that BCR/ABL expression does not affect the expression level of Bmi-1.

Thus, similar to the case in normal hematopoiesis, Bmi-1 expression is highest in immature cells and progressively decreases with differentiation in leukemic hematopoiesis. Immature KLS leukemia cells could be enriched in the Bmi-1^highlin^– population (Fig. 5E).

Bmi-1 Expression in the Myeloid Leukemia Induced by the Combination of TEL/PDGFßR and AML1/ETO
Mice transplanted with BM cells transduced with TEL/PDGFßR-ires-AML1/ETO (TP-AE) developed cachexia, severe anemia, and decreased movement at day 28–58 post-transplant (Fig. 6A). In these mice, there were occasional blast cells in peripheral blood (Fig. 6B). Mice with these findings were diagnosed with leukemia and sacrificed for analysis. Upon necroscopy, splenomegaly was observed in all diseased mice (Fig. 6B). Histological examination of the BM showed hypercellarity (Fig. 6B). The absolute number of HSCs (Flk2^– KLS) was highly elevated in mice with induced leukemia compared with mice transplanted with untransduced Bmi-1^GFP/+ BM cells (normal Bmi-1^GFP/+ mice) (65,203 ± 34,431 cells vs. 3,290 ± 729 cells, respectively; n = 3) (Fig. 6C).

We examined Bmi-1 expression levels in hematopoietic stem and progenitor cells from mice with leukemia. In the leukemic HSC (Flk2^– KLS) population, the proportion of cells expressing high levels of Bmi-1 was increased compared with mice transplanted with nontransduced bone marrow. However, Bmi-1 expression levels in leukemic HSCs never exceeded the highest level in normal HSCs. Within the Flk2⁺ KLS fraction of leukemia cells, there were two populations that were distinguished by Bmi-1 expression level, and cells with a Bmi-1^highFlk2⁺ KLS phenotype exhibited similar levels of Bmi-1 expression to Flk2^– KLS leukemia cells (Fig. 6D). Among the myeloid progenitor population of leukemia mice, the level of Bmi-1 expression was similar to that of normal mice. Similar to the CML murine model, immature KLS leukemia cells from TPAE-leukemia mice were enriched in the Bmi-1^highlin^– population.

Bmi-1 Expression Levels Can Separate Cells into Distinct Subpopulations in Nonhematopoietic Tissues
We analyzed the expression of the Bmi-1 in other tissues. We first examined the brain, liver, and testis of the Bmi-1^GFP/+ mouse at 4 weeks of age for Bmi-1 expression by immunohistochemistry. In each tissue, the majority of cells expressed Bmi-1. By microscopy, we could not readily identify distinct cell subpopulations that express high levels of Bmi-1 in the majority of tissues. However, within the liver, some cells adjacent to the portal vein appeared to express Bmi-1 strongly, although the nature of these cells is unclear (Fig. 7A).

View larger version (40K): [in this window] [in a new window]   Figure 7. Cells can be distinctly separated into subpopulations according to Bmi-1 expression in types of tissues other than the hematopoietic system. (A): GFP expression in the brain, liver, and testis of the Bmi-1GFP/+ mouse at 4 weeks of age. (B): Fluorescence-activated cell sorting analysis of Bmi-1 expression in P0 brain, postnatal liver nonparenchymal cells (4 weeks), E13.5 fetal liver, and testis (8 weeks). (C): Correlation between the Bmi-1 expression and the expression of the stem or progenitor markers that had been reported previously. Cells were separated into subpopulations according to Bmi-1 expression (GFPhigh [5%], GFPint [45%], GFPlo [45%], and GFP– [5%]) and then analyzed for the expression of the stem or progenitor markers. Abbreviations: CV, central vein; E, embryonic day; GFP, green fluorescent protein; LV, lateral ventricle; PV, portal vein; w, week(s).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
It has been already reported that Bmi-1 expression level is high in populations of immature hematopoietic cells and downregulated along with cell differentiation. In addition to this, our detailed analysis revealed that Bmi-1 expression levels are dynamically regulated within HSC-enriched KLS populations. The entire population of CD150⁺CD34^–Flk2^– LT-HSC expressed higher levels of Bmi-1 compared with short-term HSCs or multipotent progenitor cells [3, 27], suggesting that the expression level of Bmi-1 may associate with the self-renewal potential. Consistent with these results, Iwama et al. have reported that enforced expression of Bmi-1 in LT-HSCs enhances their self-renewal [10]. On the other hand, HSCs lacking Bmi-1 lost self-renewal capacity [7, 9]. Therefore, although it is still unknown how HSCs maintain Bmi-1 expression levels, it is clear that the maintenance of these high levels of expression is important for HSC function.

Previously, similar attempts were made using the other HSC-specific promoters. A GFP or zoan GFP marker was expressed under the control of the Sca-1 [28] or H2K promoter [29], and these markers were used to enrich for HSC/progenitor populations. Our Bmi-1-GFP-knock-in mice are highly promising, since there is evidence suggesting that Bmi-1 plays an important role in stem cell function. In addition, we showed that cells in other tissue types can be separated into populations based on Bmi-1 expression level. This will allow us to begin to analyze the function of Bmi-1 in these and other tissues.

In CML-like disease models, where HSC numbers were twofold elevated, Bmi-1 expression levels in leukemic HSCs were similar to those in normal HSCs, showing that upregulation of Bmi-1 does not occur in leukemogenesis by p210BCR/ABL. In the myeloid leukemia model induced by the expression of TEL/PDGFßR and AML1/ETO, where HSC numbers were highly elevated, Bmi-1 expression level in the leukemic HSCs never exceeded the highest level in normal HSCs, whereas the proportion of cells that express high-level of Bmi-1 in leukemic HSC population was increased compared with those in normal HSCs. These results suggest that Bmi-1 expression was not upregulated in the most immature LT-HSCs and that accumulation of immature HSCs, which was induced by AML1/ETO and TEL/PDGFßR expression, likely resulted in the increase in the numbers of Bmi-1^highFlk2^– KLS cells. In addition, these results suggest that the increase in Bmi-1 expression in unfractionated leukemia cells reflects the accumulation of increased numbers of immature cells rather than an increase in Bmi-1 expression on a per-cell basis.

In several models of myeloproliferative diseases and actual chronic myelogenous leukemias, the chronic phase is at the level of HSCs, whereas the blast phase or acute leukemia emerges from progeny of the chronic phase, albeit at the level of non-HSC progenitors [30–34]. We showed that leukemic KLS cells were highly enriched within the highest Bmi-1-expressing population. Several studies using human primary myeloid leukemia samples showed that most leukemic stem cells shared some of the same cell surface antigenic phenotype as normal HSCs (CD34⁺CD38^–) [35], although subsequent studies showed these to be CD90^–, not CD90⁺ as in normal HSCs [36]. The normal counterpart of these CD90^– cells appears to be a non-HSC multipotent progenitor (R. Majeti, C.Y. Park, and I.L. Weissman, manuscript in preparation). Interestingly, in the TP-AE myeloid leukemia model, a subset of Flk2⁺ KLS leukemia cells expressed Bmi-1 at very high levels: in mice, members of the Flk2⁺ subset are also non-HSC multipotent progenitors [3, 37]. It will be interesting to examine whether these Bmi-1^highFlk2⁺ KLS leukemia cells are enriched for leukemic stem cells; it will clarify whether Bmi-1 expression level is a good marker independently of normal HSC makers.

These results encouraged us to examine the hypothesis that cancer stem cells in other tissues might also be marked by high Bmi-1 expression. It has already been reported that Bmi-1 is overexpressed in many kinds of malignant tissues, including mantle cell lymphoma [38], lung cancer [39], hepatocellular carcinoma [40], brain tumors [12], and colorectal cancer [41]. Recently, Glinsky et al. reported that high Bmi-1 expression was a strong prognostic factor in several types of cancer [14]. However, it is still unknown which cells within the tumors express Bmi-1. Therefore, it will be interesting to determine whether the highest Bmi-1-expressing cells in various cancer models are enriched with cancer stem cells in the Bmi-1-reporter mice.

	CONCLUSION

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
We generated Bmi-1-GFP-knock-in mice, in which GFP was expressed under the endogenous transcriptional regulatory elements of the Bmi-1 gene. These mice allow for the isolation of viable Bmi-1-expressing cells. In normal hematopoietic cells, Bmi-1 expression is dynamically regulated, and the Bmi-1 expression level is an excellent marker for normal HSCs. In two mouse leukemia models, Bmi-1 expression was not upregulated in the leukemic HSCs, despite the increase in HSC number. Immature KLS leukemia cells were enriched within the Bmi-1^high population. Furthermore, in several other nonhematopoietic tissues, cells could be separated into distinct subpopulations with differential Bmi-1 expression. These results suggest that the Bmi-1-GFP-knock-in mice have the potential to become a useful tool for understanding the role of Bmi-1 in normal and cancer stem cells in multiple tissue types.

	DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
I.L.W. owns stock in, has acted as a consultant to, has served as an officer or member of the Board for, and has a financial interest in Cellerant and Stem Cells, Inc.

	ACKNOWLEDGMENTS

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 
The research was supported by the NIH (DK053074 and CA086065 to I.L. Weissman). N.H. was supported by a Japanese Society of Promotion of Science Fellowship, the Yamada Memorial Foundation, and the Mitsubishi Pharma Research Foundation. T.Y. was supported by a Uehara Memorial Foundation Fellowship. K.P. was supported by a Giannini Family Foundation Fellowship. We thank M. Thomasson (Washington University) for MSCV-TEL/PDGFßR-ires-AML1/ETO; W. Pear (University of Pennsylvania) for MSCV-p210bcr/abl-ires-Neo; K. Rajewsky (Harvard University) for pKSTKNeoloxP; T. Yagi (Osaka University) for pCre-PAC; G. Nolan (Stanford University) for the Phoenix virus producer cell line; I. Park (University of Michigan; now Oncomed, Inc.) for the BAC clone containing the Bmi-1 gene; and H. Zeng (Stanford University), H. Nakai (Stanford University), and T. Egawa (New York University) for technical advice; L. Jerabek for excellent laboratory management; C. Muscat for antibody preparation; and L. Hidalgo, D. Escoto, and J. Dollaga for animal care. We also appreciate D. Bryder and J. Attena for technical assistance; D. Bhattacharya, L. Allies, B. Tan, and B. Chen for critical reading of the manuscript; and all the members of Weissman laboratory for great discussion and technical help.

	REFERENCES

Top Abstract Introduction Materials and Methods Results Discussion Conclusion Disclosure of Potential... Acknowledgments References

 

1. Spangrude GJ, Heimfeld S, Weissman IL. Purification and characterization of mouse hematopoietic stem cells. Science 1988;241:58–62.[Abstract/Free Full Text]

2. Osawa M, Hanada K, Hamada H et al. Long-term lymphohematopoietic reconstitution by a single CD34-low/negative hematopoietic stem cell. Science 1996;273:242–245.[Abstract]

3. Christensen JL, Weissman IL. Flk-2 is a marker in hematopoietic stem cell differentiation: A simple method to isolate long-term stem cells. Proc Natl Acad Sci U S A 2001;98:14541–14546.[Abstract/Free Full Text]

4. Kiel MJ, Yilmaz OH, Iwashita T et al. SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell 2005;121:1109–1121.[CrossRef][Medline]

5. Roy NS, Wang S, Jiang L et al. In vitro neurogenesis by progenitor cells isolated from the adult human hippocampus. Nat Med 2000;6:271–277.[CrossRef][Medline]

6. Valk-Lingbeek ME, Bruggeman SW, van Lohuizen M. Stem cells and cancer; the polycomb connection. Cell 2004;118:409–418.[CrossRef][Medline]

7. Park IK, Qian D, Kiel M et al. Bmi-1 is required for maintenance of adult self-renewing haematopoietic stem cells. Nature 2003;423:302–305.[CrossRef][Medline]

8. Molofsky AV, Pardal R, Iwashita T et al. Bmi-1 dependence distinguishes neural stem cell self-renewal from progenitor proliferation. Nature 2003;425:962–967.[CrossRef][Medline]

9. Lessard J, Sauvageau G. Bmi-1 determines the proliferative capacity of normal and leukaemic stem cells. Nature 2003;423:255–260.[CrossRef][Medline]

10. Iwama A, Oguro H, Negishi M et al. Enhanced self-renewal of hematopoietic stem cells mediated by the polycomb gene product Bmi-1. Immunity 2004;21:843–851.[CrossRef][Medline]

11. Lessard J, Baban S, Sauvageau G. Stage-specific expression of polycomb group genes in human bone marrow cells. Blood 1998;91:1216–1224.[Abstract/Free Full Text]

12. Leung C, Lingbeek M, Shakhova O et al. Bmi1 is essential for cerebellar development and is overexpressed in human medulloblastomas. Nature 2004;428:337–341.[CrossRef][Medline]

13. Tumbar T, Guasch G, Greco V et al. Defining the epithelial stem cell niche in skin. Science 2004;303:359–363.[Abstract/Free Full Text]

14. Glinsky GV, Berezovska O, Glinskii AB. Microarray analysis identifies a death-from-cancer signature predicting therapy failure in patients with multiple types of cancer. J Clin Invest 2005;115:1503–1521.[CrossRef][Medline]

15. Overturf K, Al-Dhalimy M, Tanguay R et al. Hepatocytes corrected by gene therapy are selected in vivo in a murine model of hereditary tyrosinaemia type I. Nat Genet 1996;12:266–273.[CrossRef][Medline]

16. Kubota H, Avarbock MR, Brinster RL. Spermatogonial stem cells share some, but not all, phenotypic and functional characteristics with other stem cells. Proc Natl Acad Sci U S A 2003;100:6487–6492.[Abstract/Free Full Text]

17. van der Lugt NM, Domen J, Linders K et al. Posterior transformation, neurological abnormalities, and severe hematopoietic defects in mice with a targeted deletion of the bmi-1 proto-oncogene. Genes Dev 1994;8:757–769.[Abstract/Free Full Text]

18. Kondo M, Weissman IL, Akashi K. Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 1997;91:661–672.[CrossRef][Medline]

19. Akashi K, Traver D, Miyamoto T et al. A clonogenic common myeloid progenitor that give rise to all myeloid lineages. Nature 2000;404:193–197.[CrossRef][Medline]

20. Sawa M, Yamamoto K, Yokozawa T et al. BMI-1 is highly expressed in M0-subtype acute myeloid leukemia. Int J Hematol 2005;82:42–47.[CrossRef][Medline]

21. Mihara K, Chowdhury M, Nakaju N et al. Bmi-1 is useful as a novel molecular marker for predicting progression of myelodysplastic syndrome and patient prognosis. Blood 2006;107:305–308.[Abstract/Free Full Text]

22. Pear WS, Miller JP, Xu L et al. Efficient and rapid induction of a chronic myelogenous leukemia-like myeloproliferative disease in mice receiving p210 bcr/abl-transduced bone marrow. Blood 1998;92:3780–3792.[Abstract/Free Full Text]

23. Grisolano JL, O'Neal J, Cain J et al. An activated receptor tyrosine kinase, TEL/PDGFbetaR, cooperates with AML1/ETO to induce acute myeloid leukemia in mice. Proc Natl Acad Sci U S A 2003;100:9506–9511.[Abstract/Free Full Text]

24. Lee A, Kessler JD, Read TA et al. Isolation of neural stem cells from the postnatal cerebellum. Nat Neurosci 2005;8:723–729.[CrossRef][Medline]

25. Suzuki A, Zheng YW, Kaneko S et al. Clonal identification and characterization of self-renewing pluripotent stem cells in the developing liver. J Cell Biol 2002;156:173–184.[Abstract/Free Full Text]

26. Yoshinaga K, Nishikawa S, Ogawa M et al. Role of c-kit in mouse spermatogenesis: Identification of spermatogonia as a specific site of c-kit expression and function. Development 1991;113:689–699.[Abstract]

27. Forsberg EC, Serwold T, Kogan S et al. New evidence supporting megakaryocyte-erythrocyte potential of flk2/flt3+ multipotent hematopoietic progenitors. Cell 2006;126:415–426.[CrossRef][Medline]

28. Ma X, Robin C, Ottersbach K et al. The Ly-6A (Sca-1) GFP transgene is expressed in all adult mouse hematopoietic stem cells. STEM CELLS 2002;20:514–521.[Abstract/Free Full Text]

29. Surdez D, Kunz B, Wagers AJ et al. Simple and efficient isolation of hematopoietic stem cells from H2K-zFP transgenic mice. STEM CELLS 2005;23:1617–1625.[Abstract/Free Full Text]

30. Passegue E, Wagner EF, Weissman IL. JunB deficiency leads to a myeloproliferative disorder arising from hematopoietic stem cells. Cell 2004;119:431–443.[CrossRef][Medline]

31. Jamieson CH, Ailles LE, Dylla SJ et al. Granulocyte-macrophage progenitors as candidate leukemic stem cells in blast-crisis CML. N Engl J Med 2004;351:657–667.[Abstract/Free Full Text]

32. Miyamoto T, Weissman IL, Akashi K. AML1/ETO-expressing nonleukemic stem cells in acute myelogenous leukemia with 8:21 chromosomal translocation. Proc Natl Acad Sci U S A 2000;97:7521–7526.[Abstract/Free Full Text]

33. Traver D, Akashi K, Weissman IL et al. Mice defective in two apoptosis pathways in the myeloid lineage develop acute myeloblastic leukemia. Immunity 1998;9:47–57.[CrossRef][Medline]

34. Weissman IL. Stem cell research: Paths to cancer therapies and regenerative medicine. JAMA 2005;294:1359–1366.[Abstract/Free Full Text]

35. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med 1997;3:730–737.[CrossRef][Medline]

36. Baum CM, Weissman IL, Tsukamoto AS et al. Isolation of a candidate human hematopoietic stem-cell population. Proc Natl Acad Sci U S A 1992;89:2804–2808.[Abstract/Free Full Text]

37. Bryder D, Rossi DJ, Weissman IL. Hematopoietic stem cells: The paradigmatic tissue-specific stem cell. Am J Pathol 2006;169:338–346.[Abstract/Free Full Text]

38. Bea S, Tort F, Pinyol M et al. BMI-1 gene amplification and overexpression in hematological malignancies occur mainly in mantle cell lymphomas. Cancer Res 2001;61:2409–2412.[Abstract/Free Full Text]

39. Vonlanthen S, Heighway J, Altermatt HJ et al. The bmi-1 oncoprotein is differentially expressed in non-small cell lung cancer and correlates with INK4A-ARF locus expression. Br J Cancer 2001;84:1372–1376.[CrossRef][Medline]

40. Neo SY, Leow CK, Vega VB et al. Identification of discriminators of hepatoma by gene expression profiling using a minimal dataset approach. Hepatology 2004;39:944–953.[CrossRef][Medline]

41. Kim JH, Yoon SY, Kim CN et al. The Bmi-1 oncoprotein is overexpressed in human colorectal cancer and correlates with the reduced p16INK4a/p14ARF proteins. Cancer Lett 2004;203:217–224.[CrossRef][Medline]

This Article Abstract Full Text (PDF) All Versions of this Article: 2006-0229v1 25/7/1635    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hosen, N. Articles by Weissman, I. L. Search for Related Content PubMed PubMed Citation Articles by Hosen, N. Articles by Weissman, I. L.